performing-web-cache-poisoning-attack

SUSPICIOUS: The skill is internally consistent as an offensive security guide, but it gives an AI agent concrete exploitation procedures for web cache poisoning, cache deception, and authenticated content exposure. Install sources mentioned are mostly official and there is no obvious credential exfiltration or third-party interception, so this is not confirmed malware; however, it materially enables offensive security actions with real-world impact on target systems and other users, making it high risk.